Search results
1 – 2 of 2The purpose of this paper is to supply a numerical analysis tool to solve eddy currents induced in nonlinear materials such as steel by boundary element method (BEM), and then…
Abstract
Purpose
The purpose of this paper is to supply a numerical analysis tool to solve eddy currents induced in nonlinear materials such as steel by boundary element method (BEM), and then apply it to design and analysis of power devices.
Design/methodology/approach
Utilizing integral formulas derived on the basis of rapid attenuation of the electromagnetic fields, the paper formulates eddy currents in steel. In the formulation, nonlinear terms are regarded as virtual sources, which are improved iteratively with the electromagnetic fields on the surface. The periodic electromagnetic fields are expanded in Fourier series and each harmonic is analyzed by BEM. The surface and internal electromagnetic fields are obtained numerically one after the other until convergence by the Newton‐Raphson method.
Findings
It is confirmed that this approach gives accurate solutions with meshes much larger than the skin depth and therefore is adequate to apply to a large‐scale application.
Research limitations/implications
The eddy current is formulated by utilizing the impedance boundary condition in order to meet a large‐scale application, and so solutions near the edge are poor. In the case of better solutions being required, some modifications are necessary.
Originality/value
To lessen computer memory consumption, the parallel component of the currents to the steel surface is analyzed as a 2D problem and the normal component is obtained from the parallel component. One 2D equation for one analyzing region is discretized by dividing the region into layers adaptively and then solved. Next, another is solved sequentially. This method gives a compatible numerical analysis tool with finite element method.
Details
Keywords
Kazuhisa Ishibashi and Zoran Andjelic
The purpose of this paper is to solve generic magnetostatic problems by BEM, by studying how to use a boundary integral equation (BIE) with the double layer charge as unknown…
Abstract
Purpose
The purpose of this paper is to solve generic magnetostatic problems by BEM, by studying how to use a boundary integral equation (BIE) with the double layer charge as unknown derived from the scalar potential.
Design/methodology/approach
Since the double layer charge produces only the potential gap without disturbing the normal magnetic flux density, the field is accurately formulated even by one BIE with one unknown. Once the double layer charge is determined, Biot‐Savart's law gives easily the magnetic flux density.
Findings
The BIE using double layer charge is capable of treating robustly geometrical singularities at edges and corners. It is also capable of solving the problems with extremely high magnetic permeability.
Originality/value
The proposed BIE contains only the double layer charge while the conventional equations derived from the scalar potential contain the single and double layer charges as unknowns. In the multiply connected problems, the excitation potential in the material is derived from the magnetomotive force to represent the circulating fields due to multiply connected exciting currents.
Details